RAYMOND M. KLEIN and PATRICLA A. MCMULLEN (Eds.) Converging Methods for Understanding Reading and Dyslexia Cambridge, MA: The MIT Press, 1999, 536 pages (ISBN 0-262-11247-7, us$60, Hardcover) Reviewed by MEREDYTH DANEMAN In preface to this volume, editors Raymond Klein and Patricia McMullen promise a book that will illustrate different approaches used by scientists to understand complex skill of reading and its breakdown (p. xi). As long as readers of this volume understand that Klein and McMullen are referring to the reading of single (p. 1), and not reading of sentences or paragraphs or whole texts, they will not be disappointed by this book. Indeed, Klein and McMullen have put together an impressive collection of state-of-the-art chapters by scientists who address issue of single-word reading from a variety of standpoints and disciplines. The coverage is both extensive and intensive, and quality of contributions is uniformly high. There are chapters on normal adult reading and its (Rayner; Goswami; Levy), developmental dyslexia (Lovett; Olson, Datta, Gayan, & DeFries), acquired dyslexia (Behrmann; Buchanan, Hildebrandt, & MacKinnon; Farah), neouroimaging (Demb, Poldrack, & Gabrieli; Posner & McCandliss), and computational modeling (Plaut; Masson; Besner), as well as a final chapter that offers an integrative summary of different approaches (Carr). Rayner provides a comprehensive review of what we have learned about normal adult reading from examining readers' eye movements. He shows how eye movements have revealed important information about perceptual processes in reading, such as how much information readers acquire during a single fixation, kind of information they integrate across fixations, factors that control where they fixate. And he shows how eye movements have revealed important information about moment-- to-moment cognitive processes during reading, such as how readers process ambiguous structures and ambiguous words, how they process anaphoric terms such as pronouns, and how they make inferences. Most of what we know about single-word processing has come from studies in which words are presented in isolation (see other chapters in this volume), and so there is always concern that results might not generalize to normal everyday reading (see Daneman & Reingold, 1993). What I like about Rayner's eye movement methodology is that it allows him to study how words are processed in context of natural connected text. Goswami and Levy both focus on acquisition of reading skills. Goswami argues that phonological skills needed to discriminate onset from rime (e.g., b from eak in word beak) are very important in early stages of learning to read English, and that children use rime analogies to learn new words (e.g., use beak to learn peak and weak). If Goswami is correct, then reading instruction programs that emphasize onsets and rimes should be better than those that emphasize phonemes or whole words. Levy's training studies provide only marginal support for Goswami's position. The main message from Levy's studies is that different training methods all contribute to an important but different aspect of reading development (p. 77). For example, training that emphasizes onsets or rimes leads to faster acquisition of new reading vocabulary than does training on whole words, but training on whole words is more important for developing reading fluency. Developmental dyslexia is term applied to children who have unusual difficulty acquiring basic reading skills. The chapters by Lovett and Olson et al. reflect two rather different approaches to understanding developmental dyslexia: remediation effectiveness and behavioural-genetic analysis. Lovett describes two deficit-directed remediation programs aimed at helping disabled readers acquire reliable word identification skills - one program targets deficits in phonological processing and other program targets deficits in cognitive strategy use. …